II completed the face detection part, but in the recognition part, I still compiled the recognition separately into SO to load and post-process, but how to complete the recognition and mutual with the business of the face library, I mean I read it The videcan you tell me how to complete this business logic?o has parsed the feature vector of the face, but how do I process it with the face photo in the face library file?
how to complete this business logic?Thanks!
Hey, I’m sorry, but I really didn’t get what you are trying to do?
BTW, please share your setup with us
Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)
Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU) :GPU T4
• DeepStream Version : 5.1
• Issue Type(questions, new requirements, bugs) : Regarding the task of face recognition, I used the C++ SDK of deepstream to deploy. I have completed the analysis of the detection part and the face model. The current progress is that I use C++ to compile the code to parse the 512-dimensional vector of the face recognition arcface. , But I need to complete a completed business. There should be a database picture. After deepstream receives the stream, I need to analyze it with the locally stored picture, but Deepstream can’t read my picture for reasoning, so can I only use the model first Does my face database perform tensor storage and then compare? There is still better processing logic in deepstream？
Thanks, look forward to your reply
I think you can store the info in meta data and then do the comparison in a probe on downstream component of detector.